Macronutrient Content of Plant-Based Food Affects Growth of a Carnivorous Arthropod

Ecology, 92(2), 2011, pp. 325–332 Ó 2011 by the Ecological Society of America

Macronutrient content of plant-based food affects growth of a carnivorous arthropod

1,4 2 3 1 SHAWN M. WILDER, DAVID A. HOLWAY, ANDREW V. SUAREZ, AND MICKY D. EUBANKS 1Department of Entomology, Texas A&M University, College Station, Texas 77843 USA 2Division of Biological Sciences, University of California at San Diego, La Jolla, California 92093 USA 3Department of Entomology and Department of Animal Biology, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801 USA

Abstract. Many arthropods engage in mutualisms in which they consume plant-based foods including nectar, extrafloral nectar, and honeydew. However, relatively little is known about the manner in which the specific macronutrients in these plant-based resources affect growth, especially for carnivorous arthropods. Using a combination of laboratory and field experiments, we tested (1) how plant-based foods, together with ad libitum insect prey, affect the growth of a carnivorous ant, Solenopsis invicta, and (2) which macronutrients in these resources (i.e., carbohydrates, amino acids, or both) contribute to higher colony growth. Access to honeydew increased the production of workers and brood in experimental colonies. This growth effect appeared to be due to carbohydrates alone as colonies provided with the carbohydrate component of artificial extrafloral nectar had greater worker and brood production compared to colonies deprived of carbohydrates. Surprisingly, amino acids only had a slight interactive effect on the proportion of a colony composed of brood and negatively affected worker survival. Diet choice in the laboratory and field matched performance in the laboratory with high recruitment to carbohydrate baits and only slight recruitment to amino acids. The strong, positive effects of carbohydrates on colony growth and the low cost of producing this macronutrient for plants and hemipterans may have aided the evolution of food-for-protection mutualisms and help explain why these interactions are so common in ants. In addition, greater access to plant-based resources in the introduced range of S. invicta may help to explain the high densities achieved by this species throughout the southeastern United States. Key words: ant–hemipteran–plant mutualism; aphid honeydew; food for protection; Gossypium hirsutum; mutualisms; nectar; nutritional ecology; plant-based resources; red imported fire ant; Solenopsis invicta; Texas, USA.

INTRODUCTION interactions and, more generally, the conditions under Mutualisms are widespread among plants and animals which mutualisms are formed and maintained. and can affect population and community dynamics For some consumers, the motivation for consuming (Bronstein 1994, Wa¨ckers et al. 2005). Many mutualisms plant-based foods is clear. For example, in active animals involve the transfer of resources (e.g., plant-based such as hummingbirds and adult Lepidoptera, plant- resources such as nectar, extrafloral nectar, and honey- based foods provide a key source of carbohydrates to fuel dew) in exchange for services (e.g., pollination or the high metabolic costs of flight (Hainsworth and Wolf protection) (Bronstein 1998, Stadler and Dixon 2005, 1972, Josens and Farina 1997). However, the importance Wa¨ckers et al. 2005). Resources provided by plants or of plant-based foods for the diet of other animals is less hemipterans typically contain high concentrations of well understood. For example, the high carbohydrate and carbohydrates and low concentrations of amino acids low amino acid content of plant-based foods seems (e.g., an average of 310 g/L for carbohydrates and 3 g/L inconsistent with the high nitrogen requirements hypoth- for amino acids; Blu¨thgen et al. 2004). However, there esized for carnivorous arthropods (White 1978, Fagan et can be over 100-fold variation in the concentrations of al. 2002, Denno and Fagan 2003, Fagan and Denno carbohydrates (4–986 g/L) and amino acids (0.02–26.7 2004). Yet, carnivorous arthropods frequently engage in g/L) in plant-based resources (Blu¨thgen et al. 2004). A mutualisms with plants and Hemiptera that produce better understanding of how such variation affects the plant-based resources and consumption of these foods growth and survival of consumers will provide insight has positive effects on activity level, longevity, and growth into the costs and benefits of participating in mutualistic (Eubanks 2005, Stadler and Dixon 2005, Wa¨ckers et al. 2005, Helms and Vinson 2008). It is not known whether the effects of plant-based foods on the performance of Manuscript received 25 March 2010; revised 7 June 2010; accepted 12 July 2010. Corresponding Editor: R. A. Raguso. carnivorous arthropods are due to the carbohydrates, 4 E-mail: [email protected] amino acids, or both. Experimental tests of how the 325 326 SHAWN M. WILDER ET AL. Ecology, Vol. 92, No. 2 macronutrient content of plant-based resources affects 2009. After 48 hours, water was used to separate carnivore performance may reveal whether the low amino workers, brood, and queens from the soil. This material acid content provides more of a benefit than would be was used to make standardized laboratory colonies that expected based on its concentration or whether carbohy- each consisted of 1–20 queens (to test the effects of drates are an underappreciated macronutrient for growth aphid honeydew on colony growth) or two queens (to of carnivorous arthropods. In addition, the effects of test the effects of carbohydrates and amino acids on carbohydrates and amino acids on carnivore performance growth), ;50 brood and 1 g wet mass of workers (0.345 may provide insight into the likelihood with which 6 0.009 g dry mass or 1192 6 62 individuals; mean 6 1 individuals should engage in mutualisms with plants or SE). Polygyne colonies in the field vary in the number of hemipterans that provide different concentrations of queens present (Tschinkel 2006). However, the number these macronutrients. of queens had no effect on colony size in experiment 1 Consumption of plant-based foods may also be (general linear model ANOVA, number of workers, important for the establishment and spread of invasive F1,28 ¼ 0.26, P ¼ 0.61; number of brood, F1,28 ¼ 0.45, P ¼ arthropods. Some invasive ants may have greater access 0.51). For this reason, we used two queens in each to plant-based resources in their introduced range than colony for experiment 2. Each colony from the field was in their native range (Tillberg et al. 2007). In the used to make a single laboratory replicate. Laboratory southeastern United States, the red imported fire ant colonies were housed in containers (56 cm length 3 40 (Solenopsis invicta) is responsible for significant ecolog- cm width 3 14 cm height) lined with fluon and provided ical, health, and economic impacts (Tschinkel 2006). with a darkened petri dish lined with plaster of paris for Solenopsis invicta is a carnivorous ant that primarily a nest. Colonies were provided with a vial of water and consumes animal tissue and haemolymph and has a had the plaster of paris substrate moistened two times 15 d N stable isotope signature similar to that of obligate per week. Each colony was provided with two freshly predators, such as spiders (Tennant and Porter 1991, killed crickets, Acheta domesticus, three times per week, Vogt et al. 2002, Tillberg et al. 2007). However, S. which was ad libitum prey for colonies used in these invicta also frequently forms mutualisms with honey- experiments. The use of ad libitum insect prey is dew-producing hemipterans in their introduced range consistent with conditions in nature as colonies of S. (Helms and Vinson 2002) and these interactions invicta regularly dry and store pieces of insects in colony influence how S. invicta affects arthropod food webs chambers in the field (Gayahan and Tschinkel 2008). (Kaplan and Eubanks 2005). In addition, consumption The colony room was maintained on a 12:12 light : dark of honeydew increases colony growth rate of S. invicta, photoperiod with 40–70% humidity and a daily temper- although the mechanisms responsible for this increase ature cycle that included 8 h during daylight at 328C and (e.g., greater worker longevity, higher rates of brood 16 h at 248C. Colonies were maintained in the laboratory production, or both) remain unclear (Helms and Vinson for 45 d in experiment 1 and 60 d in experiment 2 after 2008). Testing how the macronutrient content of plant- which they were frozen. We then separated workers and based resources affects carnivorous arthropods may brood, dried them at 608C for 48 h, and weighed them. provide a better understanding of the spread, ecological impacts, and control of invasive species such as S. Effects of aphid honeydew on colony growth invicta. This laboratory experiment quantified the growth of The purpose of this study was to examine the benefits colonies of S. invicta in the presence of high and low of consuming plant-based resources for a carnivorous densities of aphids. Each container with fire ants insect, the red imported fire ant (Solenopsis invicta), and received three cotton plants, Gossypium hirsutum,in to test which nutritional components of plant-based 8.75 cm square plastic pots containing potting soil, in resources (i.e., carbohydrates, amino acids, or both) which S. invicta readily nested. The aphid treatment (n ¼ contributed to higher colony growth. In the first part of 15) was created by adding cotton aphids, Aphis gossypii, this study, we quantified the effects of aphid honeydew to plants and allowing populations to become estab- on worker and brood production of S. invicta colonies lished while the control treatment (n ¼ 17) was reared in the laboratory. In the second part of this study, maintained by searching cotton plants twice each week we separated the carbohydrate and amino acid compo- and killing any aphids found. Cotton plants had nents of artificial extrafloral nectar to test: (1) which extrafloral nectaries. However, given the low number macronutrients contributed to colony growth in the of extrafloral nectaries, nectar consumption by ants was laboratory and (2) which were preferentially collected by likely low compared to the amount of honeydew colonies in the laboratory and field. consumed (Adjei-Maafo and Wilson 1983, Fischer et

METHODS al. 2002). Cotton plants that grew too large for the plastic pots were replaced with smaller plants. Only one General methods for laboratory experiments cotton plant was replaced at a time and the pot with the Colonies of polygyne S. invicta were excavated from plant that was removed was allowed to dry for one week the campus of Texas A&M University (College Station, before it was removed so that all worker and brood Brazos County, Texas, USA) in the spring of 2008 and would move out of that pot and into the other remaining February 2011 MACRONUTRIENTS AFFECT CARNIVORE GROWTH 327 pots. At 12, 19, and 26 d after the start of the ex- This behavior allowed us to record accurate counts of periment, we counted the number of aphids on each leaf the number of dead ants without disturbing the colony. of each plant. At the end of the experiment, we removed On three separate dates (1, 9, and 16 d after the start of the soil from all of the pots and spread it in a thin layer the experiment) we also counted the number of S. invicta on the bottom of the containers. We also added a petri collecting liquid from the experimental treatment vials. dish that was painted black and included moistened We counted the number of ants on the vials 60 min after cotton to each container. As the soil in the container a new treatment vial was added to the colony. The dur- dried, the ants moved into the petri dish with the brood. ation of larval growth and pupation is ;30 d at 328Cin After 2 d, we removed the petri dish, visually searched S. invicta (Tschinkel 2006). Hence, the collection of dead containers for any remaining ants, and placed all ants ants and data on feeding from treatment vials occurred and brood in a freezer. Two-factor ANOVA was used to before colonies had a chance to increase the number of test for the effects of aphid treatment and the number of workers present. queens on the dry mass of workers and brood present in Two-factor ANOVAs were used to test for the effects colonies after 45 d on the experimental treatments. A of carbohydrates and amino acids on the dry mass of repeated-measures ANOVA was used to compare the workers, dry mass of brood, number of dead workers, mean number of aphids per leaf between plants in the and proportion of the colony composed of brood at the aphid and control treatments for the three sampling end of the experiment. A two-factor, repeated-measures dates. ANOVA was used to test for the effects of carbohydrates and amino acids on the number of ants collecting Effects of carbohydrates and amino acids liquid from the experimental treatment vials. on colony growth Diet choice in the field We conducted a second laboratory experiment to test the effects of the carbohydrate and amino acid compo- We conducted an experiment in the field to examine nents of artificial extrafloral nectar on colony growth of the preference of S. invicta colonies for solutions S. invicta. We randomly assigned colonies to four treat- containing carbohydrates, amino acids, or both. One ments that, in addition to insect prey, were supplemented potential criticism of this approach is that carbohydrates with a 5-mL vial of either: water (n ¼ 12); a solution and amino acids are present in very different concen- containing only the carbohydrate component of extra- trations and that diet selection could be dose dependent. floral nectar (n ¼ 13); a solution containing only the Dose dependence can occur in studies of diet selection; amino acid component (n ¼ 15); or a solution with both however, it is always between solutions that have dif- components (n ¼ 14). Artificial nectar mimicked the ferent concentrations of the same macronutrient (e.g., chemical composition of extrafloral nectar of Passiflora Josens and Farina 1997, Blu¨thgen and Fiedler 2004). spp. (Lanza 1991) and consisted of 1 L of water mixed When consumers are choosing between different mac- with carbohydrates (108 g sucrose, 90 g glucose, 53 g ronutrients, consumers will select the macronutrient that fructose) and amino acids (0.0232 g aspartic acid, 0.512 g is limiting and consume that macronutrient until their glutamine, 0.0404 g glutamic acid, 0.0194 g histidine, dietary requirement is fulfilled (e.g., Mayntz and Toft 0.0436 g isoleucine, 0.04 g leucine, 0.118 g phenylalanine, 2001, Mayntz et al. 2005, Raubenheimer et al. 2007). 0.368 g proline, 0.0704 g tryptophan, and 0.1122 g To facilitate counting ants feeding on different tyrosine). This artificial nectar recipe has carbohydrate solutions in the field, we combined the treatment (251 g/L) and amino acid (1.3 g/L) concentrations similar solutions (water, carbohydrates only, amino acids only, to a wide range of extrafloral nectars (carbohydrate, or both the carbohydrate and amino acid component of mean ¼ 222, median ¼ 183; amino acid, mean ¼ 3.4, extrafloral nectar) with agar to form a soft bait pellet median ¼ 1; Blu¨thgen et al. 2004). Plant-produced (112 mL of solution with 0.8 g of agar). On two dates, extrafloral nectar also contains a number of other com- we created transects of 15 bait stations spaced 10 m pounds including volatiles, used in long-distance attrac- apart and placed one of each of the bait treatments at tion, and antimicrobial chemicals (Gonza´lez-Teuber and each station (i.e., a simultaneous choice test). Transects Heil 2009). However, our artificial extrafloral nectar only were placed in post oak savannah habitat at the field included carbohydrates and amino acids because these laboratory on the campus of Texas A&M University in are thought to be the major nutritional components that College Station, Texas, in June 2009. After 60 min, we contribute to the survival and growth of consumers returned to each bait and counted the number of S. (Blu¨thgen et al. 2004, Stadler and Dixon 2005). Vials invicta feeding on the bait. Two-factor ANOVA was with experimental treatments were replaced twice each used to test for the effects of carbohydrates and amino week. acids on the number of ants at the baits after 60 min.

Twenty-eight days after the laboratory colonies were RESULTS established, we collected and counted the number of dead worker ants in each colony. Solenopsis invicta, like Effects of aphid honeydew on colony growth many other ants, places dead worker ants in one or more We were successful at creating experimental treat- piles away from the nest (Ho¨lldobler and Wilson 1990). ments with different aphid density as plants in the aphid 328 SHAWN M. WILDER ET AL. Ecology, Vol. 92, No. 2

FIG. 1. Comparison of the ﬁnal dry mass of workers and brood (mean þ SE) from colonies of the carnivorous red imported ﬁre ant, Solenopsis invicta, provided access to honeydew from cotton aphids on cotton plants (Aphid) or just cotton plants (Control). Colonies of polygyne S. invicta were excavated from the campus of Texas A&M University, College Station, Brazos County, Texas, USA.

treatment had over three times as many aphids on each between these effects (F1,33 ¼ 6.42, P ¼ 0.02; Fig. 3). leaf (130.7 6 2.7 aphids/leaf) compared to plants in the When no carbohydrates were present, the addition of control treatment (43 6 1.5 aphids/leaf) (repeated- amino acids resulted in a 57% reductioninthe measures ANOVA: treatment effect, F1,15 ¼ 32.12, P , proportion of colony mass that was brood. However, 0.001). After 45 d, colonies of S. invicta in the aphid when carbohydrates were present, the addition of amino treatment had almost 50% higher dry mass of workers (F1,28 ¼ 8.07, P ¼ 0.008) and .200% higher dry mass of brood (F1,28 ¼ 4.70, P ¼ 0.039) compared to control colonies (Fig. 1). Effects of carbohydrates and amino acids on colony growth Colonies with carbohydrate supplements had almost 100% higher dry mass of workers (F1,33 ¼ 17.59, P ¼ 0.0002; Fig. 2A) and .100% higher dry mass of brood (F1,33 ¼ 9.77, P ¼ 0.004; Fig. 2B) compared to colonies deprived of carbohydrates. There was no effect of amino acids and no interaction between carbohydrates and amino acids on the dry mass of workers (amino acids, F1,33 ¼ 0.92, P ¼ 0.34; interaction, F1,33 ¼ 0.39, P ¼ 0.53) or brood (amino acids, F1,33 ¼ 0.47, P ¼ 0.50; interaction, F1,33 ¼ 1.88, P ¼ 0.18) at the end of the experiment. Colonies with supplemental carbohydrates had 54% lower mortality (111 6 35 workers dead) compared to colonies deprived of carbohydrates (245 6 33 workers dead; F1,27 ¼ 16.37, P ¼ 0.0004). However, the consumption of amino acids resulted in a 60% increase in mortality (219 6 35 workers dead) relative to colonies that did not have amino acids (137 6 33 workers dead; F1,27 ¼ 5.40, P ¼ 0.03). While there were no significant main effects of FIG. 2. Comparisons of the effects of a full factorial manipulation of the presence and absence of the carbohydrate carbohydrates (F1,33 ¼ 2.09, P ¼ 0.16) or amino acids and amino acids components of artificial extrafloral nectar on (F1,33 ¼ 0.03, P ¼ 0.86) on the proportion of a colony the final dry mass of (A) workers and (B) brood in colonies of composed of brood, there was a significant interaction Solenopsis invicta (mean þ SE). February 2011 MACRONUTRIENTS AFFECT CARNIVORE GROWTH 329

with the supplement containing a combination of carbohydrates and amino acids had a slight increase in the ratio of brood to workers compared to colonies with only the carbohydrate supplement. The strong dietary preferences of S. invicta for carbohydrate supplements in the field matched their performance on those supplements in the laboratory. Many species of predatory arthropods consume plant-based resources in nature and show higher fitness when consuming these resources (Wa¨ckers et al. 2005). However, this is the first study to decouple the nutritional components of plant-based resources and demonstrate that carbohydrates, and not amino acids, contribute to increased production of both FIG. 3. Comparisons of the effects of a full factorial brood and workers of a primarily carnivorous ant. manipulation of the presence and absence of the carbohydrate and amino acids components of artificial extrafloral nectar on The amino acids in plant-based foods have been the proportion of a colony composed of brood (i.e., mass of hypothesized to be an important resource for arthropod brood/total colony mass; mean þ SE). mutualists and many arthropods prefer plant-based foods that contain higher concentrations of amino acids acids resulted in a 90% increase in the proportion of a (Lanza 1991, Yanoviak and Kaspari 2000, Blu¨thgen and colony that was brood (Fig. 3). Fiedler 2004, Davidson 2005). Solenopsis invicta also Workers of S. invicta recruited to vials with carbohy- shows a slight preference for amino acids. When amino drates in much higher numbers than they did to vials acids are consumed alone this does not seem to be that did not have carbohydrates (F1,26 ¼ 69.44, P , beneficial because it increases worker mortality without 0.0001; Fig. 4A). There was also higher recruitment to any benefit to colony growth. In this case, the small vials with amino acids (F1,26 ¼ 4.08, P ¼ 0.05); however, consumption of amino acids may represent some ants the effect of amino acids on recruitment (35% increase) sampling the resource or over-ingesting the overall was much smaller than the effect of carbohydrates (264% increase; Fig. 4A). There was no interaction between carbohydrates and amino acids (F1,26 ¼ 0.39, P ¼ 0.54). Diet choice in the field In a simultaneous choice experiment in the field, S. invicta recruited to baits with carbohydrates in much higher numbers than they did to vials with no carbohydrates (F1,87 ¼ 103.54, P , 0.0001; Fig. 4B). In parallel with the results in the laboratory, S. invicta recruited to vials with amino acids in higher numbers than they did to vials without amino acids (F1,87 ¼ 4.29, P ¼ 0.04), but the effect for amino acids (36% increase in recruitment) was much smaller than the effect for carbohydrates (651% increase in recruitment; Fig. 4B). There was no interaction between carbohydrates and amino acids in the recruitment of ants to baits (F1,87 ¼ 0.04, P ¼ 0.84).

DISCUSSION By decoupling the separate effects of the carbohydrate and amino acid components of extrafloral nectar, our data show that, even when provided with ad libitum insect prey, carbohydrates limited S. invicta colony growth. Colonies that had their diet of insect prey FIG. 4. Comparisons of the effects of a full factorial supplemented with carbohydrate-only supplements (ex- manipulation of the presence and absence of the carbohydrate periment 2) and honeydew (experiment 1) had similar and amino acids components of artificial extrafloral nectar on increases in worker and brood mass. Amino acid the recruitment of Solenopsis invicta workers to baits (mean þ SE): the recruitment of workers to baits (A) in the laboratory supplements had no effect on colony growth and a when they were only provided with one type of bait and (B) in negative effect on worker survival. The only potential the field when they were provided with all four bait types positive effect of amino acids was that colonies provided simultaneously. 330 SHAWN M. WILDER ET AL. Ecology, Vol. 92, No. 2 amino acid solution to gain limiting dietary essential that feeding on diets high in energetic compounds con- amino acids (e.g., five of the amino acids in our solution tributes to increased mass gain, growth, and repro- are considered dietary essential for insects; Chapman duction (Raubenheimer et al. 2007, Dussutour and 1998). However, colonies do gain a slight benefit in the Simpson 2008, 2009, Salomon et al. 2008, Cook et al. proportion of a colony composed of brood when the 2010, Kay et al. 2010). A more critical analysis of the amino acids are combined with carbohydrates. The fact nitrogen limitation hypothesis reveals that several key that the effect of amino acids is small may not be assumptions of this hypothesis may not be met, which surprising given that carnivorous S. invicta primarily may explain why it is not supported by experimental consume insects, which are rich in amino acids (e.g., data (Wilder and Eubanks 2010). While the nitrogen .60% amino acids by dry mass; Ramos-Elorduy et al. limitation hypothesis assumes that the dietary require- 1997) compared to plant-based foods (,1% amino acid ments of arthropods are directly proportional to the by dry mass; Blu¨thgen et al. 2004). Clearly, S. invicta nutrients in their bodies, the actual dietary requirements show a much higher preference and increase in colony of arthropods can be inversely related to their body growth with carbohydrates than amino acids, and this contents (Fagan et al. 2002, Wilder and Eubanks 2010). preference may have been a strong selective factor For example, in the spring, Agonum dorsale beetles have influencing the nutritional composition of plant-based high protein and low lipid contents in their bodies but foods, which typically have a lower concentration of preferentially forage for lipid in their diet to restore lipid amino acids and higher concentration of carbohydrates reserves that were depleted during diapause (Rauben- than phloem sap (Byrne and Miller 1990, Blu¨thgen et al. heimer et al. 2007). The controversy over nutrient lim- 2004, Hu et al. 2009). itation in carnivorous arthropods highlights the need for Extrafloral nectar and honeydew both contain amino direct experimental data to critically evaluate long- acids, which are thought to be costly to incorporate into standing hypotheses that have yet to be rigorously these plant-based resources (Bronstein 1998). However, tested. our data suggest that these amino acids may only have a The digestive physiology of larval and adult ants may slight benefit on colony growth of S. invicta, which contribute to the large effect of carbohydrates on colony questions why they are incorporated into plant-based growth. Most insects can metabolize carbohydrates, foods. One hypothesis is that the amino acids in plant- lipids, and amino acids to acquire energy (Paine 1971, based foods are important when insect prey are limited Wigglesworth 1984). However, in ants and some other or for species that feed less on insect prey (e.g., more social insects, castes differ in their digestive abilities ‘‘herbivorous’’ ant species; Davidson et al. 2003). Amino (Tschinkel 2006). While larvae can process a wide range acids could also act as phagostimulants and be added in of food typical of the capabilities exhibited by most extrafloral nectar by plants to distract ants from tending solitary insects, adult worker ants have a reduced hemipterans (Hanny and Elmore 1974, Becerra and digestive system that cannot process insect prey and, Venable 1989). Yet another hypothesis is that these instead, rely on carbohydrates for fuel (Vinson 1983, amino acids are not as costly as once thought. Plants Tschinkel 2006). All insect prey items are brought by and some hemipterans have mutualist microbes that workers to the larvae, processed and digested by the synthesize amino acids and there could be times when larvae, and then redistributed throughout the colony excess amino acids are produced and need to be voided (i.e., to the queen, workers, and other larvae) as needed (Fischer et al. 1998, Douglas 2006). For example, aphids through trophallaxis (Vinson 1983, Tschinkel 2006). require certain essential amino acids in their diet, which Hence, consumption of carbohydrates has a double are often limited, and will void excess nonessential benefit to colony growth of ants by directly providing amino acids in honeydew (Douglas 1993, Sandstro¨m energy to fuel worker foraging and by allowing larvae to and Moran 2001, Douglas 2006). Hence, amino acids divert more resources to their own growth and less need not be an addition designed to benefit mutualist toward processing food for workers. The dual benefit of consumers but, rather, could be an excess product of carbohydrates may explain the large effect that carbo- another mutualism between the plant or hemipteran and hydrates had on colony growth of S. invicta and why their symbiotic microbes (Fischer et al. 1998, Douglas mutualisms between ants and honeydew-producing 2006). insects or plants with extrafloral nectaries are extremely Carnivorous arthropods have long been hypothesized common in nature (Ho¨lldobler and Wilson 1990, Helms to be nitrogen limited (White 1978, Fagan et al. 2002, and Vinson 2002, Stadler and Dixon 2005). In addition, Denno and Fagan 2003, Fagan and Denno 2004). access to plant-based carbohydrates in its introduced However, recent empirical data do not support this range could play an important role in the invasion hypothesis and, instead, suggest that some carnivorous ecology of S. invicta. arthropods may be energy limited (e.g., Raubenheimer Research on nutritional ecology and ecological et al. 2007, Salomon et al. 2008). Our results and those stoichiometry has provided tremendous insight into of several recent studies have shown that some ecology and evolution at scales ranging from chemical carnivorous arthropods preferentially forage for ener- reactions to ecosystem dynamics (Sterner and Elser getic compounds (i.e., carbohydrates and lipids) and 2002, Simpson et al. 2009). However, relatively few February 2011 MACRONUTRIENTS AFFECT CARNIVORE GROWTH 331 studies have been conducted on the nutritional ecology Davidson, D. W., S. C. Cook, R. R. Snelling, and T. H. Chua. of carnivores (Raubenheimer et al. 2007) despite the 2003. Explaining the abundance of ants in lowland tropical rainforest canopies. Science 300:969–972. importance of carnivore nutrition for the ecological Denno, R. F., and W. F. Fagan. 2003. Might nitrogen impacts of predators in biological control and as limitation promote omnivory among carnivorous arthro- invasive species (e.g., Kaplan and Eubanks 2005). Our pods? Ecology 84:2522–2531. results contribute to the growing evidence that sources Douglas, A. E. 1993. The nutritional quality of phloem sap of energy (e.g., carbohydrates and lipid; Raubenheimer utilized by natural aphid populations. Ecological Entomol- ogy 18:31–38. et al. 2007, Salomon et al. 2008) can limit the growth of Douglas, A. E. 2006. Phloem-sap feeding by animals: problems carnivorous arthropods despite long-held beliefs that and solutions. Journal of Experimental Botany 57:747–754. nutrients (e.g., nitrogen and phosphorus) limit the Dussutour, A., and S. J. Simpson. 2008. Carbohydrate growth of predators and a range of other animals regulation in relation to colony growth in ants. Journal of Experimental Biology 211:2224–2232. (White 1978, Fagan et al. 2002, Sterner and Elser 2002). Dussutour, A., and S. J. Simpson. 2009. Communal nutrition in The strong benefit of carbohydrates for carnivorous ants ants. Current Biology 19:1–5. and the low cost of including carbohydrates in plant- Eubanks, M. D. 2005. Predaceous herbivores and herbivorous based foods by plants and hemipterans may have aided predators: the biology of omnivores and the ecology of omnivore–prey interactions. Pages 3–16 in P. Barbosa and I. the evolution of food-for-protection mutualisms and Castellanos, editors. The ecology of predator–prey interac- help explain why these mutualisms are so common tions. Oxford University Press, New York, New York, USA. (O’Dowd 1979, Stadler and Dixon 2005, Wa¨ckers et al. Fagan, W. F., and R. F. Denno. 2004. Stoichiometry of actual 2005). Further research on the nutritional ecology of vs. potential predator–prey interactions: insights into nitro- carnivores will aid in developing integrated models of gen limitation for arthropod predators. Ecology Letters 7: 876–883. nutrition to explain how small-scale processes such as Fagan, W. F., E. Siemann, C. Mitter, R. F. Denno, A. F. individual foraging behavior scale to community and Huberty, H. A. Woods, and J. J. Elser. 2002. Nitrogen in ecosystem dynamics (Simpson et al. 2009). insects: implications for trophic complexity and species diversification. American Naturalist 160:784–802. ACKNOWLEDGMENTS Fischer, M. K., W. Vo¨lkl, R. Schopf, and K. H. Hoffmann. We thank Kenneth Helms for comments on a previous 2002. Age-specific patterns in honeydew production and version of the manuscript. Michael Castro and Hannah honeydew composition in the aphid Metopeurum fuscoviride: McKenrick provided valuable help with laboratory and field implications for ant-attendance. Journal of Insect Physiology experiments. Funding was provided by NSF DEB 0716983 to 48:319–326. ´ M. D. Eubanks, NSF DEB 0717054 to D. A. Herre, and NSF Fischer, W. N., B. Andre, D. Rentsch, S. Krolkiewicz, M. Tegeder, K. Breitkreuz, and W. B. Frommer. 1998. Amino DEB 0716966 to A. V. Suarez. acid transport in plants. Trends in Plant Science 3:1360– LITERATURE CITED 1385. Gayahan, G. G., and W. R. Tschinkel. 2008. Fire ants, Adjei-Maafo, I. K., and L. T. Wilson. 1983. Association of Solenopsis invicta, dry and store insect pieces for later use. cotton nectar production with Heliothis punctigera (Lepi- Journal of Insect Science 8:1–8. doptera: Noctuidae) oviposition. Environmental Entomology Gonza´lez-Teuber, M., and M. Heil. 2009. Nectar chemistry is 12:1166–1170. tailored for both attraction of mutualists and protection from Becerra, J. X. I., and D. L. Venable. 1989. Extrafloral nectaries: exploiters. Plant Signaling and Behavior 4:809–813. A defense against ant–Homoptera mutualisms? Oikos 55: Hainsworth, F. R., and L. L. Wolf. 1972. Energetics of nectar 276–280. extraction in a small, high altitude, tropical hummingbird, Blu¨thgen, N., and K. Fiedler. 2004. Preferences for sugars and Selasphorus flammula. Journal of Comparative Physiology A amino acids and their conditionality in a diverse nectar- 80:377–387. feeding ant community. Journal of Animal Ecology 73:155– Hanny, B. W., and C. D. Elmore. 1974. Amino acid 166. composition of cotton nectar. Journal of Agricultural and Blu¨thgen, N., G. Gottsberger, and K. Fiedler. 2004. Sugar and Food Chemistry 22:476–478. amino acid composition of ant-attended nectar and honey- Helms, K. R., and S. B. Vinson. 2002. Widespread association dew sources from an Australian rainforest. Austral Ecology of the invasive ant Solenopsis invicta with an invasive 29:418–429. mealybug. Ecology 83:2425–2438. Bronstein, J. L. 1994. Our current understanding of mutualism. Helms, K. R., and S. B. Vinson. 2008. Plant resources and Quarterly Review of Biology 69:31–51. colony growth in an invasive ant: the importance of Bronstein, J. L. 1998. The contribution of ant–plant protection honeydew-producing hemiptera in carbohydrate transfer studies to our understanding of mutualism. Biotropica 30: across trophic levels. Environmental Entomology 37:487– 150–161. 493. Byrne, D. N., and W. B. Miller. 1990. Carbohydrate and Ho¨lldobler, B., and E. O. Wilson. 1990. The ants. Harvard amino acid composition of phloem sap and honeydew University Press, Cambridge, Massachusetts, USA. produced by Bemisia tabaci. Journal of Insect Physiology Hu, L. P., F. Z. Meng, S. H. Wang, X. L. Sui, W. Li, Y. X. Wei, 36:433–439. J. L. Sun, and Z. X. Zhang. 2009. Changes in carbohydrate Chapman, R. F. 1998. The insects: structure and function. levels and their metabolic enzymes in leaves, phloem sap and Cambridge University Press, New York, New York, USA. mesocarp during cucumber (Cucumis sativus L.) fruit Cook, S. C., M. D. Eubanks, R. E. Gold, and S. T. Behmer. development. Sciencia Horticulturae 121:131–137. 2010. Colony-level macronutrient regulation in ants: mech- Josens, R. B., and W. M. Farina. 1997. Selective choice of anisms, hoarding and associated costs. Animal Behaviour 79: sucrose concentration by the hovering hawk moth Macro- 429–437. glossum stellatarum. Journal of Insect Behavior 10:631–637. Davidson, D. W. 2005. Ecological stoichiometry of ants in a Kaplan, I., and M. D. Eubanks. 2005. Aphids alter the New World rain forest. Oecologia 142:221–231. community-wide impact of fire ants. Ecology 86:1640–1649. 332 SHAWN M. WILDER ET AL. Ecology, Vol. 92, No. 2

Kay, A. D., T. Zumbusch, J. L. Heinen, T. C. Marsh, and D. A. Stadler, B., and A. F. G. Dixon. 2005. Ecology and evolution of Holway. 2010. Nutrition and interference competition have aphid–ant-interactions. Annual Review of Ecology, Evolu- interactive effects on the behavior and performance of tion, and Systematics 36:345–372. Argentine ants. Ecology 91:57–64. Sterner, R. W., and J. J. Elser. 2002. Ecological stoichiometry: Lanza, J. 1991. Response of fire ants (Formicidae: Solenopsis the biology of elements from molecules to the biosphere. invicta and S. geminata) to artificial nectars with amino acids. Princeton University Press, Princeton, New Jersey, USA. Ecological Entomology 16:203–210. Tennant, L. E., and S. D. Porter. 1991. Comparison of the diets Mayntz, D., D. Raubenheimer, M. Salomon, S. Toft, and S. J. of two fire ant species (Hymenoptera: Formicidae): solid and Simpson. 2005. Nutrient-specific foraging in invertebrate liquid components. Journal of Entomological Science 26: predators. Science 307:111–113. 450–465. Mayntz, D., and S. Toft. 2001. Nutrient composition of the prey’s diet affects growth and survivorship of a generalist Tillberg, C. V., D. A. Holway, E. G. LeBrun, and A. V. Suarez. predator. Oecologia 127:207–213. 2007. Trophic ecology of invasive Argentine ants in their O’Dowd, D. 1979. Foliar nectar production and ant activity on native and introduced ranges. Proceedings of the National a neotropical tree, Ochroma pyramidale. Oecologia 43:185– Academy of Sciences USA 104:20856–20861. 191. Tschinkel, W. R. 2006. The fire ants. Belknap Press, Cam- Paine, R. T. 1971. The measurement and application of the bridge, Massachusetts, USA. calorie to ecological problems. Annual Review of Ecology Vinson, S. B. 1983. The physiology of the imported fire ant and Systematics 2:145–164. revisited. Florida Entomologist 66:126–139. Ramos-Elorduy, J., J. M. P. Moreno, E. E. Prado, M. A. Perez, Vogt, J. T., R. A. Grantham, E. Corbett, S. A. Rice, and R. E. J. L. Otero, and O. L. de Guevara. 1997. Nutritional value of Wright. 2002. Dietary habits of Solenopsis invicta (Hyme- edible insects from the state of Oxaca, Mexico. Journal of noptera: Formicidae) in four Oklahoma habitats. Environ- Food Composition and Analysis 10:142–157. mental Entomology 31:47–53. Raubenheimer, D., D. Mayntz, S. J. Simpson, and S. Toft. Wa¨ckers, F. L., P. C. J. van Rijn, and J. Bruin. 2005. Plant 2007. Nutrient-specific compensation following diapauses in provided food for carnivorous insects: a protective mutual- a predator: implications for intraguild predation. Ecology 88: ism and its applications. Cambridge University Press, Cam- 2598–2608. bridge, UK. Salomon, M., D. Mayntz, and Y. Lubin. 2008. Colony White, T. C. R. 1978. The importance of a relative shortage of nutrition skews reproduction in a social spider. Behavioral food in animal ecology. Oecologia 33:71–86. Ecology 19:605–611. Sandstro¨m, J. P., and N. A. Moran. 2001. Amino acid budgets Wigglesworth, V. B. 1984. Insect physiology. University Press, in three aphid species using the same host plant. Physiolog- Cambridge, UK. ical Entomology 26:202–211. Wilder, S. M., and M. D. Eubanks. 2010. Might nitrogen Simpson, S. J., D. Raubenheimer, M. A. Charleston, F. J. limitation promote omnivory among carnivorous arthro- Clissold, and the ARC-NZ Vegetation Function Network pods? Comment. Ecology 91:3114–3117. Herbivory Working Group. 2009. Modelling nutritional Yanoviak, S. P., and M. Kaspari. 2000. Community structure interactions: from individuals to communities. Trends in and the habitat templet: ants in the tropical forest canopy Ecology and Evolution 25:53–60. and litter. Oikos 89:259–266.